NTNU Norwegian University of Science and Technology Faculty of Engineering Department of Mechanical and Industrial Engineering
Ane JohnsgaardEvaluation of Asset Performance
Ane Johnsgaard
Evaluation of Asset Performance
A study of Required Function, Maintenance and Operational History
Master’s thesis in Engineering and ICT Supervisor: Per Schjølberg
Co-supervisor: Jon Martin Fordal June 2021
Master ’s thesis
Ane Johnsgaard
Evaluation of Asset Performance
A study of Required Function, Maintenance and Operational History
Master’s thesis in Engineering and ICT Supervisor: Per Schjølberg
Co-supervisor: Jon Martin Fordal June 2021
Norwegian University of Science and Technology Faculty of Engineering
Department of Mechanical and Industrial Engineering
Preface
This Master’s Thesis concludes the master’s degree in Engineering and ICT, with the main profile ICT & Operation Management at the Norwegian University of Science and Technology (NTNU) in Trondheim. It was written during the spring of 2021 for the subject TPK4950 (Reliability, availability, maintainability and safety, master thesis). This thesis is part of the Faculty of Engin- eering (IV) and the Department of Mechanical and Industrial Engineering (MTP). The thesis was written in collaboration with Hydro Aluminium and was supervised by Per Schjølberg (Associate professor, MTP) and Jon Martin Fordal (PhD Candidate, MTP) from NTNU.
Acknowledgement
While concluding this thesis, there are several that deserve to be acknowledged. Throughout this period, I have received lots of support and assistance.
I want to thank my supervisor at NTNU, Per Schjølberg, along with Hydro Aluminium, for giving me the opportunity to work on this project. Thanks to both Arnt Johnsen and Per Gullaksen for making this possible and your input during the process. Thank you to my supervisor and co-supervisor, Jon Martin Fordal, for their guidance and input.
To the employees in Hydro, I would like to thank you for answering my questions. A special thanks to Torleif Berg for checking in with me every week and helping me with what I needed. Thanks to Sondre Norhaug for answering all my questions. It is very much appreciated. Thank you to my fellow students for making these five years memorable. My family and friends deserve a thank you for all their support and motivation during these years in Trondheim. To my sister, Maiken, thank you very much for proofreading my thesis and for all your support. Thank you to my parents for all the comforting words. To Birgitte, thanks for the motivation throughout these last weeks.
Lastly, thanks to Jørgen for always listening and for all your encouragement.
Ane Johnsgaard Trondheim, 10.06.2021
Abstract
This master thesis laid the foundation for evaluating the asset performance in production facilities for the purpose of continuous improvement, decreased downtime and reduced profit loss. The performance of the assets is important for the overall business performance. Through measuring and evaluating the performance of the assets, the current situation could be better described,
”siloes” minimised, and better decisions made. By integrating the available data, it is possible to monitor the performance of the assets continuously. However, there is a gap between the theoretical concept of asset performance and the implementation in ageing production facilities.
The data available for assessing the performance is not previously indented for this task, causing the realisation of mapping the condition and current performance of a system a complex task.
To address this, a literature search and a case study were completed. The available maintenance and operational history were analysed to evaluate the required function of a transportation line for carbon anodes at Hydro Aluminium’s carbon production facility in ˚Ardal. The transportation line consists of an overhead conveyor, a roller conveyor and an automated storage, and the lack of control of the system is expanded. The transportation line transport and store anodes between the green mill, where the anodes are moduled, and the furnaces, where the anodes are baked before aluminium production. Through this analysis, including system and data analysis, critical areas of the transportation line were identified, and the effect of failures and following maintenance actions was visualised. When evaluating the required function of the system and analysing the available data, the lack of control over the system was identified, and the need for better monitoring of the system was highlighted. Through utilising condition data from the system, the function and performance of the assets can be continuously monitored.
From this mapping of the current state of the transportation line, critical areas of the transport- ation line were identified. These areas influence the total performance of the transportation line, and concrete improvement measures at these areas could raise the overall performance. From im- plementing integrated systems to monitor the current condition of the assets continuously, steps towards improved performance and condition could be made. New technological solutions would visualise the need for improvements. Future research should include how continuous monitoring would improve the performance in the long run and look further into closing the gap between the research performed and the current condition at different facilities. By integrating the divisions of a company, the overall performance will improve.
Samandrag
Denne masteroppg˚ava legg eit grunnlag for ˚a evaluere tilstanden p˚a utstyret i eit produksjonsan- legg. Dette vert gjort for ˚a jobbe med kontinuerlig forbetring, samt for ˚a redusere nedetid og minimere produksjonstap etter uventa hendingar. Tilstanden p˚a utstyret er viktig for den overor- dna prestasjonen til bedrifta. Gjennom ˚a m˚ale og vurdere tilstanden kan ein beskrive situasjonen betre, fjerne ”siloar” og fatte betre avgjerder. Ved ˚a knyte saman det dataet ein har tilgjengeleg p˚a utstyret kan ein kontinuerleg overvake tilstanden. Vidare er det ofte identifisert eit gap mellom dei teoretiske konsepta knytte til dette emnet og det ˚a setje det ut i live p˚a eksisterande og aldrande fabrikkar. Dataet som er tilgjengeleg for ˚a evaluere tilstanden er i utgangspunktet ikkje tiltenkt denne oppg˚ava. Som følge av dette vil det vere ei større utfordring ˚a f˚a kartlagt b˚ade tilstanden og ytinga p˚a utstyret.
B˚ade ein litteraturstudie og ei konkret problemstilling vart undersøkt for ˚a sj˚a nærmare p˚a denne utfordringa. Transportanlegget for karbonanodar ved Hydro Aluminium sin karbonfabrikk i ˚Ardal vart valt som eit omr˚ade ˚a sj˚a nærmare p˚a. Denne transportlinja best˚ar av b˚ade ei hengebane, ei rullebane og eit automatisert lager for ˚a frakte og lagre anodar mellom massefabrikken, som formar anodane, og anodefabrikken, som bakar anodane klare for aluminiumsproduksjon. Utfordringar knytte til at ein ikkje er klar over tilstanden og dermed manglar kontroll p˚a dette anlegget, er utbreitt. Den tilgjengelege vedlikehaldshistorikken og produksjonsdataet knytte til transportlina, vart analysert for ˚a identifisere omr˚ade av linja som er meir kritiske, samt følger av feil og p˚afølgande vedlikehald som m˚a utførast. Analysen inkluderer b˚ade ein systemevaluering og analyse av det tilgjengelege dataet. Dette vart vidare knytt til evaluering av den kravde funksjonen til anlegget, som synleggjorde mangelen p˚a kontroll av anlegget og behovet for ˚a overvake tilstanden p˚a systemet betre. Gjennom ˚a bruke tilstandsdata direkte fr˚a systemet kan b˚ade funksjonen, tilstanden og prestasjonen av utstyret bli kontinuerleg overvaka.
Fr˚a denne kartlegginga av noverande tilstand p˚a transportlinja er spesifikke omr˚ade identifisert som meir kritiske enn andre. Dette p˚averkar totalintrykket av ytinga p˚a linja, og kan truleg forbetrast med konkrete tiltak. Ved ˚a implementere integrerte system som synleggjer tilstanden p˚a utstyret kontinuerleg f˚ar ein visualisert problema, og tiltak kan setjast i verk tidlegare. Fr˚a denne kartlegginga av noverande tilstand, kan ein byggje steg for steg vidare til ei auka yting og forbetra tilstanden p˚a utstyret, samt ta i bruk nye teknologiske løysingar for ˚a synleggjere kvar forbetringar m˚a gjennomførast. Vidare forsking bør inkludere korleis kontinuerleg overvaking p˚averkar tilstanden p˚a sikt, i tillegg til ˚a sj˚a vidare p˚a korleis gapet mellom forskinga og tilstanden rundt om i fabrikkane kan minimerast. Ved at dei ulike avdelingane i bedrifta jobbar saman, kan den overordna prestasjonen bli forbetra.
Contents
Preface i
Acknowledgement i
Abstract ii
Samandrag iii
List of Abbreviations vii
List of Figures viii
List of Tables x
1 Introduction 1
1.1 Background . . . 1
1.2 Problem statement . . . 3
1.3 Research questions . . . 3
1.4 Research objectives . . . 4
1.5 Limitations . . . 5
1.6 Report structure . . . 5
2 Methodology 6 2.1 Literature search for theoretical background . . . 6
2.2 Case study . . . 7
2.2.1 System analysis . . . 7
2.2.2 Data analysis . . . 8
3 Theoretical background and previous research 10 3.1 Asset management and asset performance . . . 10
3.1.1 Asset management . . . 10
3.1.2 Asset performance . . . 11
3.1.3 Asset performance maturity . . . 12
3.1.4 The link between asset management and asset performance . . . 13
3.1.5 The connection between the condition of the assets and the asset performance 14 3.2 Asset performance and maintenance . . . 15
3.2.1 Maintenance management . . . 15
3.2.2 Maintenance processes . . . 17
3.2.3 Required function . . . 19
3.3 Digital development for asset performance evaluation . . . 19
3.3.1 Industry 4.0 and its contribution to asset performance . . . 20
3.3.2 Smart maintenance in an asset performance perspective . . . 21
3.3.3 Predictive maintenance for asset performance assessment . . . 23
3.3.4 Beyond predictive maintenance for better asset performance . . . 24
3.3.5 Condition monitoring for measuring asset performance . . . 26
3.4 Data for asset performance analytics . . . 26
3.5 Summary of the literature search and the presented theory . . . 28
4 Case study introduction 30 4.1 Norsk Hydro and Hydro Aluminium ˚Ardal . . . 30
4.2 ˚Ardal Karbon . . . 30
4.3 Digital maintenance in Hydro . . . 31
4.3.1 Ongoing project . . . 32
4.3.2 Maintenance strategy . . . 33
4.3.3 Asset performance . . . 33
4.4 Process description . . . 34
5 Case study results and analysis 36 5.1 System definition . . . 36
5.1.1 Sub-system 1 - Overhead conveyor(hengebana) . . . 38
5.1.2 Sub-system 2 - Roller conveyor(rullebana) . . . 38
5.1.3 Sub-system 3 - Automated storage(grøntlager) . . . 39
5.2 Functional and physical architecture . . . 39
5.3 Data analysis . . . 41
5.3.1 Maintenance history (SAP records). . . 41
5.3.2 Stops and downtime registrations(stopptidsregistrering) . . . 48
5.3.3 Records of production setbacks(tilbakesettingar) . . . 48
5.3.4 Combining the data from the different sources . . . 50
5.4 Critical areas of the transportation line . . . 52
5.4.1 S.23 . . . 52
5.4.2 R.11 . . . 53
5.4.3 R.20 . . . 53
5.4.4 P.3 andP.4 . . . 53
5.5 Required asset performance compared to delivered performance . . . 54
6 Discussion 57
6.1 Asset performance . . . 57
6.2 Current state of the transportation line . . . 58
6.3 Future state of the transportation line . . . 61
6.4 Asset performance for gaining back the control of the assets . . . 63
7 Conclusion 65 7.1 Further work . . . 65
References 66 Appendix 71 A Technical hierarchy from SAP . . . 71
B System analysis . . . 74
B.1 System analysis abbreviations . . . 74
B.2 System analysis figures . . . 75
C Python scripts for data analysis . . . 78
C.1 Analysis of extracted data from SAP using IW39 . . . 78
C.2 Analysis of extracted data from SAP using IW69 . . . 82
C.3 Analysis of records of production setbacks(tilbakesettingar) . . . 84
D Analysis of downtime registrations(stopptidsregistrering) . . . 87
E Analysis of SAP records . . . 91
E.1 SNUSTASJON OVN 3 POS 23(S.23) . . . 91
E.2 RULLEBANE TIL OVN 2, 3 OG 4 POS 11(R.11). . . 94
E.3 RULLBANE TIL OVN 3 OG 4 POS 20(R.20) . . . 96
E.4 PUSHROD OVN 3 POS 27(P.3). . . 98
E.5 PUSHROD OVN 4 POS 28(P.4). . . 100
List of Abbreviations
ACM asset condition management AI artificial intelligence
ALCM asset life cycle management AM asset management
AMMP asset maintenance management process ANN artificial neural network
AP asset performance
APM asset performance management
BD big data
CM condition monitoring
DM data mining
DMT Digital Maintenance Toolbox DT digital twin
ERP enterprise resource planning FIFO First In, First Out
HSE health, safety and environment IoT Internet of Things
IT information technology IV Faculty of Engineering KPI key performance indicators MES manufacturing execution system ML machine learning
MTP Department of Mechanical and Industrial En- gineering
NLP natural language processing
NTNU Norwegian University of Science and Techno- logy
OEE overall equipment effectiveness OT operations technology
PAM physical asset management PdM predictive maintenance SAM smart asset management
SCADA supervisory control and data acquisition SM smart maintenance
SMDSS smart maintenance decision support system TDL trusted data layer
TLP technical language processing VDM value-driven maintenance
List of Figures
1 Background for problem statement . . . 2
2 Relationship between research questions and research objectives . . . 4
3 Structure and progress of report . . . 5
4 Overview of the research methodology . . . 6
5 Overview of the system and data analysis . . . 9
6 Key elements of an asset management system . . . 11
7 Asset performance management maturity steps . . . 12
8 An overview of the architecture for the ACM framework . . . 14
9 Maintenance management process . . . 16
10 Value drivers in maintenance . . . 17
11 Core elements of the maintenance process . . . 18
12 Industry 4.0 development . . . 20
13 Reference architecture model Industrie 4.0 . . . 21
14 The four dimensions of smart maintenance . . . 22
15 Smart maintenance framework . . . 22
16 Steps towards predictive maintenance . . . 23
17 Functional structure for PdM . . . 24
18 Integrated framework for digital reliability and maintenance . . . 25
19 Data quality for asset performance analytics . . . 27
20 The production process . . . 30
21 Hydro’s conceptual architecture for digital maintenance . . . 31
22 Overview of Hydro’s digital maintenance development . . . 32
23 Hydro’s future maintenance process . . . 34
24 Overview of the technical hierarchy for the transportation line . . . 35
25 Overview of the transportation line . . . 36
26 Overview of the components in the transportation line . . . 37
27 Functional and physical architecture of the transportation line . . . 40
28 Number of maintenance orders per functional location . . . 42
29 Duration of maintenance orders per functional location . . . 43
30 Cost of maintenance orders per functional location . . . 44
31 Number of corrective maintenance orders per functional location . . . 45
32 Duration of corrective maintenance orders per functional location . . . 46
33 Cost of corrective maintenance orders per functional location . . . 47
34 Setbacks of the baking process per year . . . 49
35 Setbacks of the baking process grouped by categories per year . . . 50
36 The technical hierarchy for the overhead conveyor(hengebana) . . . 71
37 The technical hierarchy for the roller conveyor (rullebana) . . . 72
38 The technical hierarchy for the automated storage(grøntlager) . . . 73
39 The technical hierarchy for the pushrods . . . 73
40 Overview of sub-system 1 - The overhead conveyor(hengebana) . . . 75
41 Overview of sub-system 2 - The roller conveyor (rullebana) . . . 76
42 Overview of sub-system 3 - The automated storage(grøntlager) . . . 77
43 Registration of downtime forthe bakehouse from 2016 until 2021 . . . 87
44 Registration of stops for the bakehousefrom 2016 until 2021 . . . 87
45 Registration of downtime fortransport to furnaces from 2016 until 2021 . . . 88
46 Registration of stops for transport to furnaces from 2016 until 2021 . . . 88
47 Registration of downtime forthe green mil from 2016 until 2021 . . . 89
48 Registration of stops for the green mil from 2016 until 2021 . . . 89
49 Registration of downtime forthe overhead conveyor from 2016 until 2021 . . . 90
50 Registration of stops for the overhead conveyor from 2016 until 2021 . . . 90
51 Corrective maintenance orders per year for(S.23). . . 91
52 Components in need of corrective maintenance for(S.23) . . . 92
53 Problems leading to component breakdown for(S.23) . . . 92
54 Causes of problems leading to component breakdown for(S.23) . . . 93
55 Corrective maintenance orders per year for(R.11) . . . 94
56 Components in need of corrective maintenance for(R.11) . . . 94
57 Problems leading to component breakdown for(R.11) . . . 95
58 Causes of problems leading to component breakdown for(R.11). . . 95
59 Corrective maintenance orders per year for(R.20) . . . 96
60 Components in need of corrective maintenance for(R.20) . . . 96
61 Problems leading to component breakdown for(R.20) . . . 97
62 Causes of problems leading to component breakdown for(R.20). . . 97
63 Corrective maintenance orders per year for(P.3) . . . 98
64 Components in need of corrective maintenance for(P.3) . . . 98
65 Problems leading to component breakdown for(P.3) . . . 99
66 Causes of problems leading to component breakdown for(P.3) . . . 99
67 Corrective maintenance orders per year for(P.4) . . . 100
68 Components in need of corrective maintenance for(P.4) . . . 100
69 Problems leading to component breakdown for(P.4) . . . 101
70 Causes of problems leading to component breakdown for(P.4) . . . 101
List of Tables
1 Downtime registrations between2019-02-25 to 2019-03-03 . . . 512 Maintenance orders and notifications between2019-02-25 to 2019-03-03 . . . 51
3 The required asset performance compared to the delivered performance . . . 55
4 List of abbreviations used for the system parts . . . 74
1 Introduction
Asset performance (AP) is a relevant topic for asset-intensive companies today. With the ongoing digitalisation and the new possibilities arriving, the need to evaluate the assets’ performance has arrived. Implementation of Industry 4.0 technologies has the positive benefits of”improved pro- ductivity and asset performance, reduced inefficiencies, lower production and maintenance cost”
[1]. The Standardisation Roadmap of Predictive Maintenance for Sino-German Industry 4.0/In- telligent Manufacturing indicates that assessment and monitoring of the equipment and the assets are needed for assuring efficient production and minimising unplanned downtime, which is essential for remaining competitive [2].
A survey conducted in over seventy Swedish companies by Bokrantz, Skoogh and Ylip¨a¨a, in 2016, concluded that there is considerable potential in using engineering tools and methods during oper- ations to achieve high equipment performance. The study discovered that analysing the available data is a non-prioritised task, leading to a lack of understanding of failure occurrence and con- sequences. By prioritising this, there is a potential of identifying profit losses, which is required for increasing the performance and an essential part of the current digitalisation [3]. This is confirmed by Campos, Sharma, Jantunen, Baglee and Fumagalli while highlighting the need to improve de- cision making by basing it on the collected data. The machines’ existing data or other maintenance and production systems could improve the asset management process. The problem is that the data is not taken to use [4]. Making it possible to base maintenance decisions on the available data is also elaborated by Bumblauskas, Gemmill, Igou and Anzengruber [5].
Norsk Hydro is currently on the journey towards utilising their data for continuous improvement. In
˚Ardal, a carbon anode production facility produces aluminium for Hydro Aluminium’s aluminium facilities. Carbon anodes are necessary for the production of aluminium for conducting electricity.
Like many other ageing production facilities, there is a need to control the assets’ performance.
The product being produced has been getting the attention, and the connection between a well- functioning asset performance system and the ability to save money on maintenance and operations has been overlooked [6]. If possible, to use the available data, decisions would be made on the proper foundation and not the familiar gut feeling. A basis for asset performance evaluation could be made by analysing the transportation line’s required function and the available maintenance and operation history. When gaining control over the performance of the assets, the production efficiency could increase and the profit losses reduced.
1.1 Background
How to gain control over the assets and evaluate their performance is central to this thesis. Parida concluded already in 2012 that most industries lack a proper asset management strategy [7], and it has been better understood in the following years due to the rapid digitalisation. Sp¨untrup and Imsland stated in 2018 that different prognosis and optimisation technologies combined with an asset management strategy is crucial to improve the performance of the assets in the process industry for the future [8]. Wang, Chen and Parlikad identified the importance of connecting asset management performance and business performance, especially for asset-intensive production companies [9]. This is confirmed by Maletic et al., stating that physical asset management has a
”statistically significant impact on operational performance” [10].
Lukens, Naik, Saetia and Hu point towards the quality of the maintenance data being an obstacle for industrial companies when wanting to take their maintenance data to use for the new tech- nologies and data-analytics possibilities, to increase their overall performance [11]. Therefore, it is relevant to find ways to apply the maintenance history and operational data to evaluate the performance of the assets. New analytical methods are dependent on available data, and to be able to connect the asset performance to the business performance, the asset condition must first be found. It is indicated that in addition to the new technology, the people-relied factors [12], the management process, and the reliability-engineering provides a crucial part in enlarging the performance of the assets [13]. Thus, these aspects must also be considered when finding ways to measure the performance of the equipment.
The three-round Delphi survey with 25 maintenance experts conducted by Bokrantz, Skoogh, Ber- lin and Stahre in 2017 identified that”data analytics, interoperable information systems, big data management, emphasis on education and training, fact-based maintenance planning, new smart work procedures, and maintenance planning with a system perspective”, will influence the mainten- ance organisations by 2030 [14]. Lundgren Skoogh and Bokrantz indicated that to keep up with the digital change, including the new technologies and the more complex systems, the maintenance department must take a vital role in this process, which will reduce risks and consequences of un- planned downtime [15]. In light of this, the combination of asset performance and maintenance is essential. Bradbury, Carpizo, Gentzel, Horah and Thibert also point to the need for looking bey- ond the aspect of predictive maintenance to see how the digital and analytical tools are helpful for the entire organisation [16]. Hence, asset performance evaluation based on the available data could be seen as the first step towards incorporating the digital shift, improving the overall performance towards the assets and the maintenance organisation and overall business performance.
Maintenance is an essential part of digital change and the key to eliminate risk and reduce downtime affecting production efficiency. Condition monitoring of the assets is a course made possible by digitisation, which connects the aspects of asset performance, and maintenance [2]. De la Fuente, Gonz´alez-Prida, Crespo, G´omez and Guill´en point to asset management, performance assessment, and monitoring of systems as an essential subject in scientific literature and industrial companies.
This is due to more complex equipment and production processes, technological development and the continued focus on cost reductions, high quality, sustainability and improved safety. Greater focus towards the evaluation of the assets could give a competitive advantage [17]. The different factors creating a background for the problem statement is visualised through Figure 1.
Figure 1: Background for problem statement
In light of Hydro’s commitment to sustainability, related to the performance, the environment and climate, the local communities, the business integrity and the health and safety, the evaluation of the current performance is critical [18]. If a system can perform better, last longer, require fewer spare parts, and not hinder the production process, this will affect the sustainable footprint. There are several examples of accidents occurring due to the equipment and the maintenance, such as the latest incident in Norway, the fire at Equinor’s facility in Hammerfest [19]. This shows the need to control the asset and the processes and make sure that maintenance actions and plans align with the current condition. The Norwegian Ministry of Local Government and Modernisation points to the need for exploiting the available data as an essential resource, the possibilities of using new technologies for utilising the available data in a better way and the potential for more value through applying and sharing data throughout the value chain [20].
The issue is how to use what the scientific literature says about asset performance and adapt it to an existing production facility with complex ageing systems and present digital solutions.
Bumblauskas et al. identify a lack of scientific literature on systems where data is used for physical asset management and maintenance [5]. The Standardisation Roadmap of Predictive Mainten- ance for Sino-German Industry 4.0/Intelligent Manufacturing also indicates a gap between the theoretical and practical application of the new technologies introduced by the fourth industrial revolution [2]. Chin, Varbanov, Klemeˇs, Benjamin and Tan point to a lack of literature considering the overall life-cycle of the asset and limitations related to oversimplifying assumptions related to the lifetime and the condition of the assets [21]. The hiatus between maintenance research and industry practises is also identified by Lundgren et al. [15]. Hence, a case study utilising the maintenance and operation history and analysing the required function of a system could cut this gap while interpreting the case study results in light of the available scientific literature.
1.2 Problem statement
The purpose of this master thesis is to evaluate how monitoring of asset performance and condition is helpful for detecting areas of improvements, detecting profit loss, and improving production efficiency. The system of analysis is the transportation line between the green mill(massefabrikken) and the bakehouse (brennovn) at the carbon facility at Hydro ˚Ardal. After being moduled, the carbon anodes produced at the facility goes through the transportation line before reaching the furnace. There is no overview of the condition and performance of the transportation line. There is no way to be aware of the transportation line’s status before it causes a delay in production.
Lack of insights leads to financial loss, and improvement in this area is much needed. Through analysing the current state of the transportation line and the maintenance performed, presenting improvement measures and gaining insight into the transportation line’s performance, the thesis should present a possible future state where asset performance is crucial.
1.3 Research questions
The scope of this master thesis is prepared in collaboration with the supervisors, the representatives from Hydro, and the author of this thesis. From the problem statement, the research questions listed below are established to explore the problem.
• Question 1: How can analysis of asset performance be useful for detecting areas of improve- ment, detecting profit loss and improving production efficiency?
• Question 2: How can the transportation line of anodes from the green mill(massefabrikken) to the furnaces (brennovn) at the carbon facility at Hydro ˚Ardal be improved with asset performance analysis?
1.4 Research objectives
The problem statement of the master thesis is divided into research objectives that correspond to the research questions. The goal is to answer the research question by performing the research objectives and the problem elaborated. Researching and presenting relevant theories regarding the topics are a crucial part of the thesis. The theory will be used as a basis for the case study presented by Hydro. Meetings and discussions with different employees in Hydro will be an essential part of the work with the case study, together with the author’s earlier experience from working at the production facility. Analysis of relevant data regarding the area of study is crucial for getting an understanding of the situation. The research objectives are listed below.
• Objective 1: Discuss how asset performance and condition monitoring can affect productiv- ity, efficiency and minimise profit loss
• Objective 2: Present the current state of the transportation line, related to production, maintenance and costs
• Objective 3: Analyse how the transportation line meets the required function
• Objective 4: Identify areas of improvement and their effect on the production, the profit losses and the asset performance
• Objective 5: Present the future state for getting better insight into the transportation line’s performance
All objectives will be necessary for both research questions. The first objective will answer the first research question. In combination with the first objective, the second and third objective build the foundation for the fourth objective. This leads up to the fifth and final objective, which answers the second research question. The relationship between the research questions and the research objectives is shown in Figure 2.
Figure 2: Relationship between research questions and research objectives
For the first objective, a literature review and presentation of the relevant theory is crucial. Ques- tions as to what and how to measure performance and the usefulness of the results are essential.
Information related to the production facility, the system for evaluation and maintenance and op- erations information is relevant for the second objective. To reach the third objective of the thesis, both system analysis and an analysis of relevant data of the transportation line is essential. The transportation line needs to be defined, and the architecture of the transportation line must be analysed. The relationship between required and delivered function is to be found. For the fourth objective, data analysis of the available data is necessary for gaining insight into the actual status of the transportation line. Based on the system analysis and the data analysis, the fifth objective aims to answer how to get a better insight into the transportation line.
1.5 Limitations
Some limitations could affect the result of this master thesis. The scope is limited to the research questions and research objectives presented in Section 1.3 and Section 1.4. The thesis is tied to the case study from Hydro. The results found are therefore not directly transferable to other situations and companies. The methods and relevant theory are, on the other hand, universal and is therefore useful in other situations. The problem and research objectives for this thesis revolve around capturing the transportation line’s current status analysed to say something about the performance of the assets. Suggestions and possible improvements based on the current status and the theoretical background are presented. It would be beyond the scope of the thesis to investigate the details and implementation of a continuous asset performance monitoring system.
The development of such a system could be based on the results highlighted by analysing the current status of the transportation line. This goes beyond the intention of this thesis.
The period for this thesis is the spring semester of 2021. The project start was set to 15.01.2021, while the delivery was set to 10.06.2021. This gave a 20 weeks long project period, and the thesis execution was limited to this period. The ongoing Covid-19 pandemic affected this master thesis largely. Due to the virus, physical communication was limited. The project meetings had to be completed digitally. The virus also made it impossible to perform company visits at the industry partner. For most of the thesis period, it was recommended to work from home.
1.6 Report structure
This master thesis consists of the sections listed below, as well as this introductory section and the appendix.
• Section 2: Methodology
• Section 3: Theoretical background
• Section 4: Case study introduction
• Section 5: Case study results and analysis
• Section 6: Discussion
• Section 7: Conclusion
Section 2 describes the methodology of the thesis. Both a literature review for gaining the theor- etical background and a case study will be conducted. Section 3 presents the relevant theory and Section 4 gives an introduction to the case study. The results and analysis of the transportation line investigated are found in Section 5. The case study results are discussed in light of the theor- etical background in Section 6, before the conclusion of the thesis is presented in Section 7. Lastly, the references and appendix are found. In the appendix, the extensive results from the analysis of the transportation line and the scripts for the data analysis are included. The overall report flow is shown in Figure 3.
Figure 3: Structure and progress of report
2 Methodology
A literature review and a case study were completed to address the problem of this master thesis.
This section answers how the research questions and research objectives were addressed and de- scribed how data was gathered, analysed, and quality assured. Both qualitative and quantitative methods were used. This triangulation of methods, where qualitative and quantitative methods are combined, strengthens the results and makes it possible for the two directions to fulfil each other. The topic of asset performance was chosen due to the increased focus on the subject from the industry partner. The case study was selected due to this particular system being a part of the production facility associated with significant consequences due to breakdowns, in addition to being an area where no previous improvement project was initiated. An overview of the research methodology is given in Figure 4. The actions were supported by information from discussions and talks with employees in Hydro and the author’s background knowledge from working experience in Hydro. The theoretical background gained from the literature review is essential for interpreting the results from the following system analysis and data analysis as part of the case study. The following sections describe the different methods further.
Figure 4: Overview of the research methodology
2.1 Literature search for theoretical background
A literature search was conducted to create a theoretical background of relevant aspects of the problem. This background was necessary for identifying improvements and a future state for the case study. Since performing a literature search is a qualitative method, where interpretation is based on the literature itself, it was necessary to consider credibility, transferability, and conform- ation. A literature review is an essential research method to keep up with the current research and works as a foundation for all types of research. The literature then serves as the background for future ideas and directions in the field [22].
The literature search was conducted through different databases for scientific literature and Oria, NTNU’s collections. ScienceDirect, Google Scholar, Scopus and Web of Science have all been used to gather relevant literature. The focus was on newer literature from the last couple of years to restrict the search results and ensure that the newest information on the topics was included, as this is vital due to the industry’s current rapid change and development. The chosen academic literature and research papers were required to be peer-reviewed to ensure high quality and validity.
In addition to academic literature, papers from different actors established in the industry were included to ensure that the current situation in the industry was considered. A breach between the theoretical solutions and the status around in different production facilities is identified, and the degree of digitalisation and technological development does not always correspond [2]. Therefore, it is vital to include the industry’s point of view for better insight into the current situation in production facilities. These articles capture the issues, and current status in the industry and have therefore been necessary for the topics of this report. The different standards published on the topics was also an essential part of the theoretical background of this thesis.
When performing the literature search, several search words were used. These were both combined and filtered with logical operators, such asand,or andnot. The search was narrowed to the five latest years. In addition to these searches, references and cited articles from articles that fulfilled the search criteria were searched. Search words used were among others: asset performance, asset management, asset performance management, asset analysis, maintenance, maintenance analysis, maintenance history, smart maintenance, digital maintenance, condition monitoring, industry, process industry, profit loss andsystem analysis.
2.2 Case study
The second part of this thesis, the case study, had the transportation line from the green mill to the bakehouse at the carbon facility at Hydro ˚Ardal as an area of analysis. This study was performed to analyse the transportation line itself and analyse available data associated with the line. This information was supported with information about the transportation line and area from Hydro’s internal systems, discussions with Hydro employees, and information gathered during the period the author worked at the facility.
Using a case study as a research method was necessary to include the complexity of the chosen case, including planning, management, analysis, and writing. The key to case studies is to understand the case analysed, and therefore it was essential to keep in mind that a case study is unique, and therefore generalisations are hardly found. Through the case study, better insight into a topic may be found [23]. Case studies are often used when questions as”how” and ”why” needs to be answered, as the case was in this situation. The type of research data used varies between the different studies, but it is essential to see the analysed data in context to understand the case overall. The research method allows gathering lots of data for gaining a thorough insight into the case chosen [24].
2.2.1 System analysis
To understand the system for examination, the transportation line, a model of the system had to be established. Visualisation of the transportation line was crucial for realising the connections between this system and the rest of the facility and the different components in the transportation line. A model or a visualisation of the transportation line makes it easier to answer the questions about the system. System block diagrams help to understand the role of the different assets [25].
The International Standard ISO 15288:2015 defines system engineering as an ”interdisciplinary approach governing the total technical and managerial effort required to transform a set of stake- holder needs, expectations, and constraint into a solution and to support that solution throughout its life” [26], and the concept is relevant when studying a complex system for evaluation of its performance. In line with this standard, it is necessary to identify assumptions, perform system analysis, review the results and use this with the recorded results as a basis for recommendations [26]. When preparing for the system analysis, it was essential to identify and define problems, questions, stakeholders, scope, objectives, methods, strategy and plan for the system analysis [26].
The needed data for the analysis could then be collected.
The standard NORSOK Z-008 (Risk-based maintenance and consequence classification) perform consequence classification in line with ten steps. This includes investigating the technical in- formation and input from other analysis, which is used to identify the system’s main function and sub-functions. Redundancy and consequences are then assigned both for the main and sub- functions. The final steps include mapping the system’s function to the equipment executing this function [27]. The same steps were necessary when analysing the system of this case study.
An architectural framework of the system would establish”a common practice for creating, inter- preting, analysing and using architecture descriptions”[28]. For visualising the architecture of the system, a combination of different diagrams was taken to use. An analysis of the required function of the system was also a part of the system analysis. Different parameters were then analysed using a table utilised in Hydro. The table is developed to understand the function of the different systems better when working with improvement measures.
To understand the relationship between the system’s functional architecture and physical archi- tecture, both the functions and components of the system had to be identified. When analysing the function, it was essential to identify the functions provided by the system, while the physical architecture should answer how the system can perform those functions [29].
The data used for the system analysis was descriptions of the transportation line and the com- ponents from Hydro’s internal systems. The technical hierarchy from Hydro’s enterprise resource planning (ERP) system,SAP PM, was used to identify the different components in the transport- ation line. These technical hierarchies are shown in Appendix A. The data analysis supported the system analysis and architecture definition, bringing crucial information for the required function analysis. By understanding the transportation line, creating diagrams of the systems, analysing the functional and physical architecture, and analysing the required function of the transportation line compared to the given performance, the system analysis was completed.
2.2.2 Data analysis
The process of data analysis started with deciding on what to analyse and how to perform the analysis. In light of the problem statement and the case study, the question of what was already explored. All the available data had to be analysed to support the system analysis. The data had to be collected and cleaned before the visualisation of the data in combination with the system analysis could be interpreted in light of the theoretical background. By analysing the available data related to the transportation line, decisions could be made on relevant information, instead of the typical gut-feeling [25].
The data used for the data analysis was obtained fromSAP PM [30], where maintenance records of all the maintenance orders completed on the transportation line were recorded for over the last twenty years. Information such as cost of the repairs, duration of the repairs, description of problem, priority, and cause related to each functional location is found. The data was collected from the system in excel sheets. First, the technical locations of the different components in the transportation line had to be obtained, and through different searches in the system, the wanted data and needed information were acquired. Secondly, the data then had to be cleaned and systematised. The focus of the analysis was limited to the last ten years. This was performed through short scripts written in the programming languagePython [31], where the data analytics library pandas [32] and the mathematical library NumPy [33] was crucial. For visualisation of the data, the visualisation library inPython, Matplotlib [34] was used. The scripts for the data visualisation are found in Appendix C, and further details about the analysis and the results of the analysis are found in Section 5 and Appendix E. Additional data for support and to identify production losses was also obtained and analysed in the same way. Downtime and stop registrations was extracted from Hydro’s manufacturing execution system (MES) system, APICS. Records of production setbacks were located from an excel sheet used for recording this. Introductory, the plan was also to include data from the supervisory control and data acquisition (SCADA) systems for measuring the performance, but the appropriate data was not available. The details about the data analysis are further described in Section 5.3.
Figure 5: Overview of the system and data analysis
To summarise the methodology of this paper, both a literature search and a case study was performed. The case study was completed through several steps, visualised in Figure 5. An understanding of the transportation line was established by analysing the technical hierarchies, building system diagrams, analysing the functional and physical architecture, and finally looking further into the required performance and function of the transportation line. The data was then captured, cleaned, transformed, and visualised to map the transportation line’s current situation and make it possible to identify problem areas and possible improvements in light of the literature search. The literature search and review results are presented in the next section before the case study, and its results are elaborated.
3 Theoretical background and previous research
As described by Wang, the asset performance (AP) will in asset-intensive production companies directly determine the performance of the business [9]. It is therefore essential to have a clear overview of the performance of the assets. A literature search was conducted to gain a theoretical background and an understanding of the relevant concepts, theories, and methods. This theoretical background is needed for combining different theories and as a basis for interpreting the results of the case study. Definitions of the important terms are included, as well as previous research and possible solutions. This chapter presents the concept of AP, looks closer to digital development and its contribution to AP and connects the concepts of AP and maintenance.
3.1 Asset management and asset performance
Asset management and AP are essential for the future. Through the theoretical model presented by Lima, McMahon and Costa, the link between asset management and business performance is elaborated. Knowledge of how investments in assets could lead to better business performance caused by the increased asset performance is the reason for the increased focus on asset management in the industry. The direct link between overall company performance to investment is essential for companies needing reasons behind their investments. If the right information is presented for decision-making, linking the goals of the company to the potentially increased asset performance, a better decision could be made [35]. In the following sections, asset management and asset performance will be further defined before the connection between them are drawn.
3.1.1 Asset management
Asset management (AM) is from the International Standard ISO 55000:2014 defined as the
”coordinated activity of an organisation to realise value from assets” and includes the task of finding the right balance of”cost, risks, opportunities and performance benefits” [36]. Anassetis defined as”an item, thing or entity that has potential or actual value to an organisation”[36]. The benefits of AM include, among others, improved financial performance, informed decision making, risk managing, and improved sustainability, efficiency, and effectiveness. The fundamentals of the concept are value, alignment, leadership, and assurance. An asset management system is described to include a context of the organisation, leadership, planning, support, operation, performance evaluation and improvement. The standard defines that ”the organisation should evaluate the performance of its assets, its asset management and its asset management system”[36].
Data management, data transformation, monitoring, analysis and continuous evaluation of asset data is presented as necessary for AM. By monitoring the performance of the assets, improvement measures will be directly identified. Relevant activities for obtaining asset information is data management, condition monitoring (CM), systems engineering, value management and availability, reliability and maintenance support [36]. The goal of AM is described as”to enable an organisation to realise value from its assets as it pursues its organisational objectives”[25]. While working for an asset management strategy, the risk of getting communication issues between the different divisions in a company, such as operations and maintenance or between maintenance and management, is reduced. To understand the condition and the value of the assets in an organisation, all the different elements of the organisations must work together [25]. The process and elements of an asset management system are described in Figure 6, where plans, policies, objectives, implementation and the assets themselves lead to the performance evaluation, which identifies AM as the foundation for further asset performance.
A study published in 2020 examines how the physical asset management (PAM) practises influence operational performance through analysing survey data from 138 organisations. The core practises of PAM was identified to be strategy and planning, risk management, life-cycle delivery, asset information and asset review. A conceptual framework linking PAM and operational performance was developed. The study shows that the PAM practices have a”statistically significant impact”
on operational performance and that companies benefit from focusing on asset management. It
is also shown that AM affects the company’s sustainability positively, both in the short and the long term, due to both the performance and the operations becoming more sustainable [10]. By focusing on the assets, more value could be extracted throughout the lifetime. These findings link AM with asset performance, leading up to the definition of AP in the next section.
Figure 6: Key elements of an asset management system
Source: Adapted from [36]
3.1.2 Asset performance
ISO 55000 definesperformanceas”measurable results”, and in the context of AM, performance is related to”assets in their ability to fulfil requirements or objectives” [36]. In the International Standard ISO 55001:2014, the performance evaluation is further described. It states that an organisation must determine what and when to monitor and measure, which methods to use and when to analyse and evaluate the results from the performance evaluation. The standard also elaborates on establishing criteria for the required processes and gaining control over the process in line with these criteria. By gaining control over the process and monitoring the assets, both corrective and preventive actions related to failures in asset performance could be performed to work for continuous improvement related to both sustainability and effectiveness [37]. As a result of this, AP is a central part of an improvement process, and the details related to how to incorporate it into the organisations must be solved.
Parida confirms that”performance needs to be measured for managing technical asset throughout its entire life cycle” [7]. Assessment of asset performance is crucial for a company’s economic and business aspects, and more companies utilise measurement of their asset performance in their business objectives and strategies. The article reflects on several issues related to asset performance.
It is a complex issue with several factors, and there is a lack of integration between”stakeholders and their changing requirement in strategic performance assessment”[7]. How to assess the performance of the assets is also an essential aspect of the issue. Operation and maintenance cost are noteworthy
in asset-intensive industries, making it more important to keep track of investments on assets to reach business requirements [7].
In the International Standard ISO 55002:2018, AP is set as an objective for AM. It is crucial to monitor the assets that could affect the value of the organisation. Both the technical, operational and financial aspects are important when evaluating performance. The decision-making methods, given criteria, risks, and opportunities are vital when deciding upon what, when and how to monitor assets. Performance monitoring could be used to identify patterns, provide needed information for decision-making, setting performance metrics, and require the needed information about the accurate status of the assets. When working towards an AP activity, it is important to identify failures and failure modes of the assets, provide sufficient information, look into detectability, research the historical evidence, and investigate cots, benefits, time horizon, and risks [38]. Thus, the AP process should start with the available historical data to map out the details needed for measuring the actual asset performance.
3.1.3 Asset performance maturity
Deacher, Das, Dunn and Sniderman describe an asset performance management (APM) program that goes beyond maintenance. It makes it possible to integrate several other business aspects, making it feasible to”optimise operations and safety, and drive financial results” [6]. A program as described would integrate the data across the companies. Both information technology (IT) and operations technology (OT) would bring critical information to the table, making it possible to draw complex decisions. Today, many asset-heavy companies cannot see the connection between having an excellent program for asset performance and savings in both maintenance and operations. The program developed has six steps of maturity, described in Figure 7. As progressing through the maturity steps, more data could be integrated into the system, enabling several functions. The report presents several examples of how sharing information between several divisions and actors in the organisation would help get better insight into the current situation, and hence, make it possible to work toward the exact directions when no longer having opposite goals. An integrated asset performance management system makes it possible to see how small changes would have severe effects and deliver holistic views of both reliability and safety [6]. The maturity steps of AP is in line with the digital development, as described throughout Section 3.3.
Combining the different departments of a company, such as operations and maintenance, motivate AP as a step towards the future.
Figure 7: Asset performance management maturity steps
Source: Adapted from [6]
3.1.4 The link between asset management and asset performance
The European Standard EN 16646:2014 defines physical asset management as ”the optimal life cycle management of physical assets to sustainably achieve the stated business objectives”and focuses on the value the assets provide, not the asset itself [39]. Physical assetis defined as”item that has potential or actual value to an organisation”, by the European Standard EN 13306:2017 [40]. By evaluating the performance of the assets, the standard points to the risks of ”silo”
behaviour, lack of holistic picture, making the wrong decisions, and uncertainty in decision-making could be avoided [39].
While delivering a framework for how value-based asset management could be implemented, Roda, Parlikad, Macchi and Garetti state that because of the central role AM has while developing strategies, the value of a companies assets must be measured to be used as a basis for decisions.
The article describes value as involving”balancing costs, risks, opportunities and benefits arising from the way assets are specified, procedures, deployed, used, maintained and disposed”[41], which exemplifies how vital it is to understand the value of the assets, making it possible to make informed decisions based on the correct information. In ISO 55000, it is outlined that asset creating value for the companies is a crucial element of AM [36]. By implementing an asset management strategy, while also making sure that the communications between the organisations in the company, the value should be obtained from the assets [42]. Volkova and Kornienko describe the importance of developing asset management strategies for different units of a company to make it possible to monitor both the different units, in addition to the entire company. It is vital to make sure that the company’s diversity is shined to light, instead of drowning in unified indicators [43]. To make sure that the value of each asset is taken into account when making decisions, the division of asset monitoring throughout the company is therefore essential.
Schuman and Brent presented back in 2005 anasset life cycle management (ALCM)model directed towards assets of the process industry. The important factors for the different stages of the life cycle of the assets are included. The goal was to optimise the value extracted from the assets in a process plant during the entire life cycle. Recommendations based on this model was to integrate asset management early on when planning new facilities. Both maintenance and operational factors should be addressed. From the model, several performance measures from the utilisation stage of the assets are mentioned. For human reliability, the model points to continuous improvement of culture and root cause failure analysis. Monitoring of the process and the assets is mentioned to improve the process reliability. Optimisations strategies and making use of history is important for equipment reliability. These measures are important to improve the overall asset performance [44]. The digital development from 2005 until today is powerful, but the same ideas presented back then are also crucial today. Process and asset monitoring and historical analysis are seen as the solution to many of the challenges faced in the process industry today. How the digital development affect the possibilities are elaborated in Section 3.3.
Wan suggest that reliability could be maximised in power grids through asset performance man- agement. The ongoing digitalisation makes it possible to overcome ”silos” in organisations to ensure that analysis of the available data is performed. Among other business drivers, the ageing assets worldwide increase the need for decisions concerning both operation and maintenance. In a case study in an electrical company, the neglect of asset inspections and maintenance and the lack of asset condition data were replaced with an asset performance system where all the condition monitoring data was utilised. This lead to a 50% reduction in asset failures. An asset model where all asset data is included and knowledge about the operations, the equipment, the data solution needs, and the industrial communication make it possible to perform better decisions regarding asset management, improving the overall performance of the assets [45]. Rødseth, Strandhagen and Schjølberg also identify the challenges regarding ”silos” where poor collaboration between the organisations in a company leads a sub-optimised production. By integrating both maintenance and operation departments, the collaboration and coordination between the departments should be improved and lead to a better understanding of how the departments work together [46]. To integrate the company’s divisions, extract the value of the assets, make decisions on a better found- ation and monitor the assets, the concepts of AM and AP are linked. The following section will review how the condition of the assets and the AP are connected to be capable of fulfilling the ideas mentioned.
3.1.5 The connection between the condition of the assets and the asset performance As Wilson describes, it is vital to find the connection between the condition and the performance of the assets. The assets is described as an”integrated package”because of the connection between the different assets. All the assets in an area or organisation influence each other highly, and together they deliver a production facility’s overall performance. However, it is essential to identify the connections between the overall system and every part it consists of. To identify failure modes and how the performance is affected, the system must either be evaluated from the top-down point of view, from facility, system, sub-system until asset, or from failure to cause [47]. An understanding of the system function and components must be established to identify how the condition affects the asset performance.
Downtime in a production facility has several consequences, such as loss of production, recover- able and not, financial loss, loss of goodwill from customers and the need for spare capacity [25].
Rødseth, Schjølberg, Kirknes and Bernhardsen present the”hidden factory”to identify profit loss and waste in production. Profit loss is defined as the total turnover loss and extra costs [48]. To prevent unscheduled downtime, Diaz-Elsayed, Hernandez, Rajamani and Weiss present a frame- work for asset condition management (ACM), which will improve the assets by delivering real-time condition monitoring, diagnostics of the assets and predictions for the future, since a manufacturing company depends on high system performance for meeting the production demand.
The framework has three cores, as presented in Figure 8. The idea is that the useful rest lifetime of the assets and the different production systems should be increased and the sustainable impact of less waste and reduced resources. The framework has six levels, or capability levels as they are called: ”limited failure indicators, diagnostics, asset monitoring, prognostics, comprehensive ACM and self-adaptive ACM” [49]. The goal is to gain control over the condition and performance of the asset, not to reach the highest level. A lower capability level could be sufficient as it depends on the given situation [49]. This framework connects asset condition, asset performance and asset management into one strategy, where the available data is used to hinder unscheduled downtime, waste and profit loss.
Figure 8: An overview of the architecture for the ACM framework
Source: Adapted from [49]
In the study by Scarpellini, Testa, Magoni and Riva, it is shown that the performance model developed can assess the reliability of different components. It is pointed out that different di- gital asset management methods are crucial to assess the status and health of the assets. Data analytical methods are therefore seen as the foundation for asset management and asset mainten- ance strategies. Healthy equipment is needed to prevent process downtime, and by monitoring the asset, it is easier to make sure that the equipment stays healthy. Statistical data, sensor data,
environmental data, and operational data were taken to use in the study’s model. The study confirmed that the operational parameters had a significant impact on the health of the assets, and a model monitoring the assets as proposed is shown crucial for an asset management solution [50]. This shows the importance of utilising the available data to monitor the assets for continuous performance evaluation, which will affect the maintenance of the assets. The following section will therefore look closer into the connection between exactly AP and maintenance before the effects of the digital development to AP is elaborated in Section 3.3.
3.2 Asset performance and maintenance
The relationship between asset management, asset performance and maintenance needs to be fur- ther investigated. EN 16646 focuses on the relationships between maintenance and events through the life cycle of the assets. There are relationships between maintenance and operations, mainten- ance and modernisation, maintenance and disposal, maintenance and physical asset management supports, and maintenance and management of assets. From managing assets, maintenance gets in- formation about physical asset management and maintenance organisations structure, objectives, policies, strategies, methods, procedures, and control systems. The asset management process gains information about life cycle cost, the impact of maintenance strategies and activities from the maintenance process [39]. This shows that maintenance decisions affect the entire life cycle of the assets and the current performance and status of the assets throughout the life cycle. Hence, the bond between maintenance and asset performance is very significant. In the following sections, maintenance management, maintenance processes and required function will be described in light of asset performance.
3.2.1 Maintenance management
The European Standard EN 13306:2017 definesmaintenanceas”the combination of all technical, administrative, and managerial actions during the life cycle of an item intended to retain it in, or restore it to, a state in which it can perform the required function”[40]. Divisions of maintenance are corrective maintenance, defined as ”maintenance carried out after fault recognition and intended to restore an item into a state in which it can perform a required function” [40], pre- ventive maintenance, defined as”maintenance carried out intended to assess and/or to mitigate degradation and reduce the probability of failure of an item” [40] and improvements. Corrective maintenance is again divided into planned and unplanned, while preventive maintenance is split into predetermined and condition-based [40]. The different maintenance performed will affect the current status of the assets, and therefore, the maintenance management and the process around the maintenance decisions and actions are essential for the performance of the assets.
Chin et al. present the state-of-the-art maintenance management for chemical process industries for asset maintenance optimisation. This sector mainly depends on the performance and condition of the assets due to the complex systems and need for stable production. There is a need for performing the correct maintenance at the right time. The importance of maintenance in the industry has risen and is often a large portion of the overall budget [21]. This show how influential the maintenance actions are toward the performance of the asset, and hence, the maintenance management must be in order.
The maintenance management process visualised in Figure 9 shows the different aspects of the maintenance process. The work process, the results and the available resources are all covered [27]. The technical condition is highly relevant for the asset performance, and from Figure 9, the following actions of reporting, analysis and improvements are thereby strongly linked. The actions leading up to the technical conditions, including goals and requirements, maintenance programme, planning and executing of the maintenance, are also necessary when assessing the asset performance. Every decision leading up to the current technical condition will affect the status and performance, thus exemplifying how strong the relationship between maintenance and asset performance is.
Figure 9: Maintenance management process
Source: Adapted from [27]
Okoh, Schjølberg and Wilson show how maintenance influences all different life-cycle phases of the assets, from product development to decommissioning. This shows that maintenance is a crucial part of asset life and, therefore, both the assets’ performance and the management of the assets.
To extract the value of the assets, maintenance needs to be performed to make sure that the assets can provide value. Through presenting the asset maintenance management process (AMMP), maintenance and asset management are integrated. Maintenance management could be improved through the asset management system, ensuring maintenance also contributes to better asset performance. Through poor or lack of maintenance, companies have experienced losses from production downtime and harm to both the assets, humans and the environment. When connecting asset management with maintenance management, the asset’s value through the entire life-cycle could be optimised [51]. Through an asset management strategy, where the asset performance is measured, the value of the maintenance actions is visualised.
As argued by Marais and Saleh, the value of maintenance is often overlooked when trying to minimise the cost of maintenance performed while increasing reliability. The article concludes that every maintenance strategy should include an aspect of the value of maintenance [52], which is also described in Haarman’s concept of value-driven maintenance (VDM). The idea is that through the four value drivers of maintenance: asset utilisation, resource allocation, cost control and health, safety and environment (HSE), maintenance could bring significant value [53, 54].
These value drivers are visualised in Figure 10. The purpose is to find the ideal balance between the four value drivers. This is important when looking further into maintenance processes in the following section since every maintenance action’s value during the process affects both the overall value and performance of each asset.
Figure 10: Value drivers in maintenance
Source: Adapted from [54]
3.2.2 Maintenance processes
The European Standard EN 17007:2017 gives an overview ofmaintenance processesand the as- sociated indicators for these processes. When managing maintenance, key activities are related to establishing policies, strategies and development actions, budgets, overseeing actions, communicat- ing crucial information, and defining areas of improvement. The core elements of the maintenance process are visualised in Figure 11. Characterisation of undesirable events is also necessary for evaluating the performance of assets. To do so, EN 17007 states key activities in this process to determine the primary cause and effect of this cause and prioritise maintenance actions based on the cause and its effects. When it comes to the process of managing data, which is a crucial part of evaluating the performance of the assets and defining areas for improvement, the standard defines the purpose of this process to”collect, analyse, store and transmit all data needed to document and improve the maintenance process” [55]. The key activities are related to storing the data, evalu- ating the reliability and maintainability of the items by assessing the state of the items, drawing up a list of necessary items, evaluate and analyse the available data, compare maintenance prac- tices, monitor methods and technologies, and finally provide access to performance and monitoring indicators [55].
